"In case anyone might find it useful..."



  • That_Annoying_KidThat_Annoying_Kid Sire of Titles Join Date: 2003-03-01 Member: 14175Members, Constellation
  • juicejuice Join Date: 2003-01-28 Member: 12886Members, Constellation
    edited September 2010
    HL1 certainly did render over 100 fps for me with dev 1 on a 160 hz monitor vsync on. I remember testing fpsmax 120 versus 160 and there was a clear difference by visualizing it with measurement of frame-to-frame-object-displacement with constant high turn speed. Of course you need to compare for the object at the same angle on the screen (like the center crosshair), but it produces clear results that you can even discriminate with the naked eye.

    <!--quoteo(post=1566447:date=Sep 16 2006, 07:34 AM:name=Jabba_The_Hunt)--><div class='quotetop'>QUOTE (Jabba_The_Hunt @ Sep 16 2006, 07:34 AM) <a href="index.php?act=findpost&pid=1566447"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->That is not at all unusual considering the age of the half life engine, one thing you should note is that you can not see more frames than your monitors refresh rate, some people believe having a higher frame rate still makes gameplay seem "smoother" even when your monitor isn't displaying the frames and I would suggest its the above type of effects which lead to this.<!--QuoteEnd--></div><!--QuoteEEnd-->

    With vsync off, you can see more frames, up to as many as you can render, even if some are delayed until the monitor refreshes and they cover some of the earlier frames. This still allows you to get part of the screen from a more recent state. So it isn't true that you can't see more frames than your monitors refresh rate. You just can't have a total number of complete frames more than your monitors refresh rate.

    Also, fps is an average. There is a great deal of variation in time per frame such that 300 fps can seem smoother than 160 fps, if only for the reason that some funky game state causes, for example, a 20 ms render time for a frame every 97th frame, even if all other frames take only 5 ms. In this extreme example, you're getting 194 fps average, up to a polling of twice a second. But twice per second, you get a delay equivalent to a gross 50 fps, which is actually noticeable when turning or moving quickly. So that's about a factor of 4. If you have 10 ms per frame constant, that's 100 fps of awesome, but when you dip to 50 ms for a frame, that's 20 fps of lame.

    This is kind of getting off topic at this point, but it does emphasize the importance of stable framerates, not just high fps, which does seem to be a factor in NS2 graphics optimization... i guess if it doesn't impact the physics like in ns1, it's not as big of a deal

    Neato stats btw, civilian. I remember first playing NS1 with 14 fps and jp was more like a parachute, so it's fun to look at the details and "troughs" and such
Sign In or Register to comment.